The purpose of this study is to examine the effect of child care quality on children's socioemotional development. Day care quality will be operationalized through variations in three state's child care regulations (Massachusetts-high quality; Virginia-moderate quality; and Georgia-low quality), by observations of caregiver-child interaction, and through center demographics (e.g., staff-child ratio) and caregiver qualifications (e.g., education and experience). Child care quality will also bb assessed as a function of setting type: (1) for profit (KinderCare), (2) non-profit/church sponsored, (3) non-profit/not church sponsored, and (4) family day care homes. The focus is on socioemotional development, because there is agreement among developmentalists that child care does not affect intellectual development and debate that child care may be a risk factor for attachment, aggression, and perhaps other indicators of socioemotional adjustment. There are five main aims to this comprehensive study of child care. The first is to determine whether legislatable quality indicators are associated with other indexes of quality. We hypothesize that they are. The second is to determine whether quality indicators will be associated with socioemotional development. We hypothesize that quality. indicators that affect caregiver-child interaction(e.g., class size and caregiver education) are most likely to be important. However, we hypothesize that these relationships will be moderated by the family system. Thus, the fourth aim is to assess interdependence between the family and child care systems. For example, we predict that associations between child care type and attachment will be moderated by family stress and maternal separation anxiety. The fourth aim is to examine caregivers as subjects and determine how their attitudes and perceptions about child care affect quality. The fifth and most important aim is theoretical: to examine the role of experience in socioemotional development. We argue that this research will add not only to our social policy base, but also to our theoretical base, because child care environments, unlike family environments, do not necessarily confound experience and genes.